DISCRIMINATION AND COMPARISON. 407 



tliem under distinct heads. It often happens, when the 

 interval is long between two experiences, that our judg- 

 ments are guided, not so much by a positive image or copy 

 of the earlier one, as by our recollection of certain facts 

 about it. Thus I know that the sunshine to-day is less 

 bright than on a certain day last week, because I then said 

 it was quite dazzling, a remark I should not now care to 

 make. Or I know myself to feel better now than I was last 

 summer, because I can now psychologize, and then I could 

 not. We are constantly busy comparing feelings with 

 whose quality our imagination has no sort of acquaintance 

 at the time — pleasures, or jiains, for example. It is notori- 

 ously hard to conjure up in imagination a lively image of 

 either of these classes of feeling. The associationists may 

 prate of an idea of pleasure being a pleasant idea, of an 

 idea of pain being a 23ainful one, but the unsophisticated 

 sense of mankind is against them, agreeing with Homer 

 that the memory of griefs when past may be a joy, and Avith 

 Dante that there is no greater sorrow than, in misery, to 

 recollect one's happier time. 



Feelings remembered in this imperfect way wvst be 

 compared with present or recent feelings by the aid of what 

 we know about them. We identify the remote experience 

 in such a case by conceiving it. The most perfect way of 

 conceiving it is by defining it in terms of some standard 

 scale. If I know the thermometer to stand at zero to-day 

 and to have stood at 32° last Sunday, I know to-day to be 

 colder, and I know just how much colder, than it was last 

 Sunday. If I know that a certain note was c, and that this 

 note is d, I know that this note must be the higher of the 

 two. 



The inference that two things differ because their con- 

 comitants, effects, names, kinds, or — to put it generally — 

 their signs, differ, is of course susceptible of unlimited 

 complication. The sciences furnish examples, in the way 

 in which men are led, by noticing differences in effects, to 

 assume new hypothetical causes, differing from any known 

 heretofore. But no matter how many may be the steps by 

 which such inferential discriminations are made, they oil 

 end in a direct inUiition of difference someiuhere. The last 



